Data Scientist : Charter Communications
Company: Shiftcode Analytics, Inc
Location: Stamford
Posted on: March 16, 2023
|
|
Job Description:
3-6 month contract to hire
Visa : USC/GC/GC-EAD
at least 3 years of strong (US based) SQL experience and 3 years
(US Based) SAS/python and predictive modeling experience. need
someone who not only understand the data but also understands the
business impact of these models and the data/insights.
someone coming out of financial industry would be preferred but not
required. someone who has built predictive models for a company in
financial industry for example.
project focus: they are building a predictive model to evaluate
credit risk in real time for potential new customers. they want to
assess risk in real time so they can make financing offers to them
as they look to sign up
they currently have a model in production also that they need to
keep checking performance on. doing performance tracking and make
adjustments to the model accordingly based on the
tracking/findings.
they are expecting the delinquency rate to rise this year due to
the economic downturn. their models are not equipped to factor this
in so they must do an analysis / forecasting based on expected
economic effects this year.
JOB SUMMARY
The Data Scientist will build predictive and statistical models to
improve credit policy and minimize credit losses. The position
leads development and analysis of new Credit Services strategies,
partnering with Credit Services subject matter experts, as well as
operations partners, to create new approaches and improve risk
mitigation performance. The successful candidate will be
responsible for generating the quantitative analysis necessary for
data-driven business recommendations.
MAJOR DUTIES AND RESPONSIBILITIES
REQUIRED QUALIFICATIONS
Skills/Abilities and Knowledge
Education
Bachelor's degree in computer science, statistics and/or equivalent
combination of education and experience. Preferred quantitative
master in science.
Related Work Experience Number of Years
Experience with SAS, Python for statistical model building 3+
Writing queries using SQL language 3+
some questions from Shang Zhao for recruiters to ask Data Scientist
candidates:
background: their team will start out with over 100 variables but
they cannot use all of them because it would take years to develop
a model of this size. they typically narrow it down to about 15-20
variables they actually use.
Q: how would you go about narrowing down this list of variables for
your model?
Q: how do you evaluate your model's predictive power in the
future?
Q: how do you validate your model's predcitve power?
Keywords: Shiftcode Analytics, Inc, Stamford , Data Scientist : Charter Communications, Other , Stamford, Connecticut
Click
here to apply!
|